Nuclease-mediated depletion biases in ribosome footprint profiling libraries

Figures

• 

  FIGURE 1.. rRNA depletion and fragment length perturbation by commercial kits. (A) Simplified flowchart of library preparation for ribosome profiling, with steps at which various rRNA depletion strategies were performed indicated. Arrowhead on circularized cDNA indicates footprint 5′ end. (B) Fractions of reads that are too short, unmappable, or mapping to various genomic regions for different rRNA depletion methods in HEK293T cells. (C) Length distribution of ribosome footprints mapping to coding regions. (D,E) Same as B and C for two different ranges of fragment size selection, depleted with the NEB kit in K562 cells. (F,G) Same as B and C, for various depletions strategies in K562 cells. In panel G, only Ribo-Zero plus data sets are shown, as others did not have changes in length distribution.

• 

  FIGURE 2.. Effect of nuclease-mediated rRNA depletion on gene-level expression measurements. Correlations between log10 CDS-mapping RPM (reads per million) for replicates, and for libraries depleted with (A) NEBNext or (B) Ribo-Zero plus (C) legacy Ribo-Zero and riboCop. Pearson correlation of log-transformed values is indicated. Fragments of all lengths were included for computing RPMs for all panels.

• 

  FIGURE 3.. Average gene profiles for depleted data sets. Average gene plots aligned at start codons for libraries depleted with (A) NEBNext or (B) Ribo-Zero plus. Reads have been offset to indicate ribosomal P sites. Reads are normalized by the read density of their respective coding region, such that all genes with sufficient read numbers contribute equally, regardless of expression level. The Ribo-Zero O, F, and OF data sets have very low numbers of CDS-mapping reads, leading to low-quality average gene plots.

• 

  FIGURE 4.. rRNA depletion with targeted biotinylated oligos. (A) Fractions of reads that are too short, unmappable, or mapping to various genomic regions for antisense oligo rRNA depletion in HEK293T cells. (B) Length distribution of ribosome footprints mapping to coding regions. (C) Average gene plots aligned at start codons for undepleted libraries or oligo depleted libraries, broken up by read length. (D) Correlations between CDS-mapping RPMs for undepleted libraries or oligo depleted libraries.